The Road Accident Analyzer. Part III: Manual and case-study E313. This report explains the functioning of a GIS-tool, called the Road Accident Analyzer (RAA), which was developed to identify dangerous road segments on the road network. In fact, the report is a manual for the use of the RAA, in which the necessary input data, the process and the results are explained stepwise for each methodology by means of a case-study of the E313. The different methodologies used by the RAA to identify dangerous road segments are the following: - Accident density; - Accident risks; - The Empirical Bayes approach. Each of these methodologies has some advantages and disadvantages. The first method for example, is very straightforward and easy to use, but it neglects some important explanatory variables for the occurrence of traffic accidents (such as traffic volume). The second method however is capable to include those road characteristics in the analysis, but assumes that the number of accidents and traffic volume are linearly related. The third method is considered as the state-of-the-art, bit is also more complex in comparison with the other methodologies. Next to the manual for the use of each methodology, these advantages and disadvantages are also being listed in this report. Finally, a sensitivity analysis on the weights which are used to include injury severity into the analysis is also carried out. 4 scenarios, similar to previous research within the Policy Research Centre of Mobility and Public Works – track Traffic Safety, are being investigated: - 1_1_1 in which each accident is evenly important; - The weights that are currently used in Flanders (1_3_5); - 1_1_10 in which the focus is on deadly accidents; - 1_10_10 in which mainly accidents with fatally and severely wounded victims should be avoided. (Author/publisher)
Abstract